Contents XIII 5 Soft Tissue Tumours in Children Gina Allen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 5.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 5.2 US . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 5.2.1 Plain Radiographs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 5.2.2 Magnetic Resonance Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 5.3 Computed Tomography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 5.4 Nuclear Medicine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 5.5 Disease characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 5.5.1 Benign Lesions Seen On US. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 5.5.2 Vascular Anomalies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 5.6 Potential Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 References and Further Reading. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 6 Interventional Techniques David Wilson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 6.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 6.2 Biopsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 6.2.1 Soft Tissue Masses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 6.2.2 Bone Masses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 6.3 Aspiration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 6.4 Local Anaesthetic Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 6.5 Osteoid Osteoma Ablation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 References and Further Reading. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 List of Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Congenital and Developmental Disorders 1 1 Congenital and Developmental Disorders David Wilson and Ruth Cheung D. Wilson, FRCP, FRCR R. Cheung, FRCR Department of Radiology, Nuffi eld Orthopaedic Centre, NHS Trust, Windmill Road, Headington, Oxford, OX3 7LD, UK CONTENTS 1.1 Introduction 1 1.2 Developmental Dysplasia of the Hip 1 1.2.1 Clinical Background 1 1.2.2 Role of Imaging in Detection 2 1.2.2.1 US Methods 3 1.2.3 Role of Imaging in Treatment 4 1.2.4 Potential Developments 5 1.3 Focal Defects 7 1.3.1 Clinical Background 7 1.3.2 Role of Imaging 8 1.3.3 Potential Developments 9 1.4 Talipes Equinovarus 9 1.4.1 Clinical Background 9 1.4.2 Role of Imaging 9 1.4.3 Potential Developments 10 1.5 Neural Tube Defects 10 1.5.1 Clinical Background 10 1.5.2 Role of Imaging 11 1.5.3 Potential Developments 16 References and Further Reading 16 1.1 Introduction There are a large number of congenital birth defects that affect the spine and appendicular skeleton. They range from isolated defects affecting one part of the body to complex syndromes with several body sys- tems involved. In practice most patients are atypical and some may cynically suggest that each case is a new syndrome. However, there are real reasons for giving as accurate a description as possible. Prog- nosis and outcome may be predictable and there is likely to be concern about the type of inheritance. Geneticists will look for as precise a diagnosis as possible and radiology, especially plain films, is part of that process (Fig. 1.1). The dysplasias that predominantly involve the skeleton may be classified in a variety of ways, but the commonest is to define the part of bone most affected, epiphysis, metaphysis or diaphysis. Sub- groups include the region of the skeleton most affected or other nonskeletal disorders. For exam- ple, spondyloepiphyseal dysplasia is a condition that affects the epiphyses and the spine. There are a good number of texts that compre- hensively describe syndromes that affect the muscu- loskeletal system and the reader is referred to them for the analysis of a particular case. In this chapter, we deal with those disorders where the imaging has a particular pivotal role in management and where ultrasound (US) has a special value. 1.2 Developmental Dysplasia of the Hip 1.2.1 Clinical Background Developmental dysplasia of the hip (DDH) is a diagnosis made when the infant’s hip is either abnormally shallow or even dislocated at birth but also when a shallow hip fails to mature to one that is mechanically stable. Its cause is not fully understood. Although there is a genetic predispo- sition, there is also evidence that abnormal stress on the hip in the later stages of pregnancy may lead to modelling deformity [1]. If untreated, a full dislocation will lead to the child failing to walk normally at around one year of age. A shallow and potentially unstable hip may not cause any symp- toms until much later in life when the abnormal stresses lead to an acetabular labral tear or pre- mature osteoarthritis. DDH diagnosed in infancy, by clinical examination and plain film analysis, is reported to occur between one and three times per thousand live births; the incidence of shallow or dysplastic acetabulae is much more frequent 2 D. Wilson and R. Cheung [2]. It is difficult to identify statistics to support this comment, but experience suggests that per- sisting shallow acetabulae are at least ten times more common. Whilst many of these children will remodel and spontaneously recover stability, some will fail to mature properly and require a variety of complex surgical procedures [3]. It has been argued that around one-tenth of hip replacements are per- formed for premature osteoarthritis secondary to mild or subclinical hip dysplasia. The goals of diagnosis and treatment are to permit affected children to walk normally and to prevent premature degeneration. We consider detection and treatment separately. 1.2.2 Role of Imaging in Detection Most developed countries have established clinical screening methods to detect children with dislo- cated or dislocatable hips and there are advocates of this as the sole screening test [4]. The manoeuvres of Ortolani and Barlow are effective in detecting around 74% of cases of dislocation or subluxation that may be demonstrated on imaging. The level of training and experience required to accurately perform these tests is substantial, and sadly the task is often placed in the hands of the more junior members of the team. There are undoubtedly occa- sions when a child with DDH is overlooked when a clinical abnormality might have been detected by a more experienced clinician. Training and audit of practice are crucial, but even in the best hands there will be errors, as clinical manoeuvres alone are not capable of detecting every case. Indeed it is also likely that some stable hips become unstable, and if the timing of the clinical examination does not coincide with this developing problem then a child may miss the chance of early treatment that could potentially limit or reverse the process. The need for early diagnosis is based on the window of opportunity that exists in the first few months of life when relatively simple treatment may be very effective. Methods range from wearing double nappies to splint therapy and corrective sur- gery. In general the later the diagnosis is made the harder the treatment will be, leading to greater risk of complications and a higher chance of failure [5]. There is a real need for a method of diagnosis that is simple, cheap, safe and effective, and US arguably provides such a technique. Unfortunately, the prac- tice of US screening for DDH has developed with no randomized control trials to judge its efficacy, and the only evidence is from observational studies, albeit with very large numbers of cases [6]. In early infancy plain films will not show the fem- oral head or much of the acetabulum as these struc- tures are not ossified until later in the first year of life. Whenever reasonable, plain film examination should be deferred until 3 to 6 months of age when more structures are ossified. Radiographs will dem- onstrate malalignment of the hips and show anoma- Fig. 1.1 Plain fi lm of the forearm of a child with metaphyseal chondrodysplasia. This examination is part of a full skeletal survey. Congenital and Developmental Disorders 3 lies of the pelvis and sacrum. The initial plain film examination should be performed without any gonad shielding as this normally overlaps parts of the pelvic ring and sacrum. Defects in these areas such as sacral agenesis may otherwise be masked. Subsequent examination should use the shields to minimize radiation dose. Despite these comments, subtle or even moderate degrees of acetabular dys- plasia will not be seen on plain films, especially in early infancy when treatment is more effective. CT and MRI would be effective ways of exam- ining the cartilaginous parts of the hips and they would allow assessment of the three-dimensional shape of the acetabulum. However, the high radia- tion burden from CT and the need for anaesthesia or sedation for most infants undergoing MRI preclude these as practical screening methods. US is safe, rel- atively cheap and repeatable with no need to sedate the infant. Its disadvantages are that it is labour- intensive and it requires skill and specific training both to perform and interpret the images. Studies have shown great sensitivity for US and a number of national bodies now require routine US screen- ing of infants for hip dysplasia. Others, including those of the United Kingdom, recommend that US is used only in infants at high risk of developing DDH (Table 1.1). Table 1.1 Risk factors for DDH Female (not a criterion commonly used in high-risk screening protocols) First degree relative with hip dysplasia Premature birth Breech presentation Other congenital limb defects Spinal defects 1.2.2.1 US Methods 1.2.2.1.1 Morphology The method pioneered and developed by Reinhart Graf in Austria has gained the widest acceptance [7]. The infant is examined shortly after birth or at least in the first 6 weeks. The infant is laid in a foam-lined trough in the lateral decubitus position. The knee and hip of the uppermost side are flexed. The US probe is placed in a true coronal plane over the hip and the angle adjusted to give an image that shows the maximum depth of the acetabulum (Fig. 1.2). Fig. 1.2 US examination of the hip images the cartilaginous structures that are invisible on plain fi lms in a coronal plane. 4 D. Wilson and R. Cheung Care must be taken not to place the probe at an oblique angle to the coronal plane as the hip may be made to look erroneously deep or shallow. The need for a precise plane of imaging is a critical issue that demands training and audit of the technique. Measurements are made from the US image to assess the amount of the bony and cartilaginous cover of the femoral head by the acetabulum either using angles or the Morin (Terjesen) method in which the proportion of femoral head lying within the cavity is measured (Fig. 1.3) [8, 9], or the Graf technique (Fig. 1.4). Hips that are shallow in comparison to the normal population are reassessed at an interval of 1 or 2 weeks and if there is failure to develop normal acetabular cover then splint therapy is commenced. Immediate therapy is started without a follow-up study when the child has already reached an age where the opportunity to treat would be lost. 1.2.2.1.2 Dynamic Examination Whilst there is some evidence that treatment may be based solely on the shape of the acetabulum, others argue that subluxation is a dynamic process and using the real-time capabilities of US it is possible to detect abnormal movement predicting dysplasia with perhaps greater sensitivity. The methods used vary but in general they are modifications of the stress tests of Ortolani and Barlow combined with US examination [10]. Gentle but firm pressure is placed on the upper part of the leg as if to subluxate the hip in a posterior and/or lateral direction. Move- ments of as little as 1 mm may be detected. How much movement is normal is contentious but some argue that over 2 mm of displacement on light stress is significant and requires treatment. It is probably wise to use both static and dynamic assessment in each case. 1.2.3 Role of Imaging in Treatment The rate of splint therapy varies with individual prac- tices and is said to be higher in those medical envi- ronments where strict conformity with treatment for abnormal US grading of acetabular dysplasia is applied. Fig. 1.3 US of the hip in the coronal plane with lines drawn to measure the amount of the head confi ned within the acetabu- lum (Morin/Terjesen method). A ratio of the overall width of the head is used as reference. Congenital and Developmental Disorders 5 Alternatively, it is argued that US screening may allow safe reduction in the numbers treated [11, 12]. Once an abnormal hip has been detected (Fig. 1.5) and treatment established there is a need to follow progress both of the shape of the acetabulum and the maturation of the bone. Over-aggressive manipulation and splint therapy may damage the growing epiphysis which will lead to deformity and delay in ossification. The latter is seen best on plain films or MRI. A reason- able approach is to repeat the US examination at follow- up appointments every 2 to 4 weeks during splint ther- apy [13] and then to perform a plain radiograph at the end of treatment or at 3 months of age (Fig. 1.6) [14]. Delay in ossification of the shallow side is expected but osteonecrosis will show much more severe retardation and then fragmentation. If there is doubt an MR study with coronal and axial T1- and T2-weighted images will detect or exclude femoral head necrosis. When surgery is required to relocate a dislocated hip then imaging with an axial cross-section technique (CT or MRI) is important to ensure correct reduction [15–17]. Frontal view plain films may easily lead to pos- terior dislocation being overlooked. Lateral plain films are usually uninterpretable in a child with the hips in a plaster spica. The child is usually sedated and quiet immediately after surgery and the limbs are held in a cast; it is therefore relatively simple to acquire cross- sectional images. MR is the preferred technique to avoid radiation, although CT is equally effective (Fig. 1.7). Planning of corrective osteotomies will require careful imaging. A combination of plain films, CT with thin low-dose sections and reconstruction, and MRI may be required [18]. Measurements may be taken from the workstation. Surface 3D reconstruc- tion images are sometimes an aid to the surgeon. Newer software algorithms that give semitranspar- ent images from multislice CT are especially useful as they mimic plane films and are better appreciated by those undertaking surgery. 1.2.4 Potential Developments US examination is playing a greater role in the monitoring of suspect dysplastic hips [19] and will Fig. 1.4 US of the hip in the coronal plane with lines drawn to measure the Graf angles. A table of measurements is used to classify the shape of the hip [7]. 6 D. Wilson and R. Cheung Fig. 1.5 US of a hip that is severely sub- luxed and almost dislocated. The “egg” of the femoral head is not sitting in the “spoon” of the acetabulum. The ace- tabular cartilage labrum is echogenic (bright), a sign seen when the tissue is stressed mechanically. Fig. 1.6 The plain fi lm appearances of the infant with hip subluxation seen at 3 months of age. The right femoral capital epiphysis has not ossifi ed and the femur is aligned in a shortened and laterally placed position. The acetabu- lum is very shallow. Congenital and Developmental Disorders 7 increasingly be used to determine the type and dura- tion of treatment [20]. It is likely that our under- standing of how and when to treat will advance as we use US to study outcome of therapy. One area of contention is whether early treat- ment by splint therapy is effective. Large numbers of infants have been the subject of routine US screen- ing and US-guided therapy in central European countries. Early data suggest that the incidence of late presentation dislocation of the hip may be much lower if not abolished [21, 22]. It will be interesting to see what happens to the rates of hip replacement in adults in the same population. Doppler US or MRI with intravenous contrast agents has been advocated as a means of predicting osteonecrosis of the treated hip. Technically these are difficult examinations and these methods have not gained wide acceptance. It might be argued that once the damage to the vascular supply has occurred there is little that can be done to reverse the process, and the treatment will be salvage of what remains of the femoral head when the repair processes are complete. Universal screening of all infants for DDH using US may seem a sensible approach but there is no con- sensus that this is reasonable at present [7, 23–28]. Not least is the doubt that splint treatment is neces- sary in all abnormal cases [29]. National policies on screening will in part reflect these awaited outcome studies but they may also be influenced by resources and health-care funding [30, 31]. The research will have to stand up to strict scrutiny before govern- ments are likely to release the substantial funds required to establish universal US screening for DDH [32]. 1.3 Focal Defects 1.3.1 Clinical Background Apart from systemic disorders or syndromes there are infrequent cases of congenital limb deficiency or malformation. Thalidomide-associated phocomelia is the best known of this type of lesion (Fig. 1.8). Sporadic cases of unknown cause are the most fre- quent now that greater care is taken over prescribing any drugs during pregnancy. Focal defects include missing bones, absent joints, single forearm or lower Fig. 1.7 Axial MRI immediately after surgical reduction of a dislocated hip with the pelvis in a plaster spica. The right femoral head is small and the acetabulum shallow but they are now properly aligned. 8 D. Wilson and R. Cheung leg bones, and absence of a segment in a dermato- mal pattern. There are sometimes associated abnor- malities of other systems, e.g. Holt-Oram syndrome where radial deficiency in the forearm is associated with a cardiac lesion. Defects of limb formation are now often recog- nized during pregnancy particularly at the 20-week “anomaly screening” examination [33–35]. It is common for the paediatric orthopaedic surgeon to be asked for advice on how such lesions might be treated by parents anticipating the need of their unborn child. 1.3.2 Role of Imaging Imaging will be required to define the extent of the defect, predict progressive deformity that may occur during maturation and to plan surgical correction. In general the key is to define the anatomy as well as possible. Technically this is often very difficult. The infants are small and they move. The bone is not yet ossified and the structures involved are very abnormal in shape. A combination of imaging will be required. Plain films are a prerequisite. They should be taken in planes as close to frontal and lateral as pos- sible. Complex projections tend to confuse. MRI is very effective but the best surface coils and thinnest sections should be used. Conforming to true sagit- tal, coronal and axial planes will help. Conventional spin echo images are probably the easiest to interpret. Cortical bone will be of low signal on all sequences and difficult to see. Cartilage gives high signal on T2-weighted images. In the immature skeleton it is difficult to differentiate unossified cartilage from adjacent soft tissue. In a deformed limb the pattern and age of ossification is variable and unpredict- able. A combination of CT and MR is useful as the bones are much better seen on CT and the cartilage is easiest to discriminate on T2-weighted images. US is very effective in showing unossified cartilage and the dynamic element allows the examiner to bend joints and demonstrate whether there is an intact joint or potential joint in an unossified carti- lage block. US is most productive if performed after Fig. 1.8 The hand and vestigial upper limb of a child with phocomelia. Congenital and Developmental Disorders 9 plain film and cross-sectional examinations. The examiner should have all previous imaging to hand before the US study. Rarely, contrast agents may be needed to demonstrate joint spaces; these may be introduced by needles guided by US and then imaged by fluoroscopy. 1.3.3 Potential Developments Improved resolution of MR and US equipment will be invaluable in assessing these complex cases. The optimum timing for surgery and therefore imaging is not always clear and as experience increases this question may be answered. 1.4 Talipes Equinovarus 1.4.1 Clinical Background Club foot is a condition of unknown cause, although it has been noted that the incidence is increased fourfold after amniocentesis. In some cases there is an association with a neurological defect, but there are also genetic and perhaps vascular factors [36, 37]. It usually presents at birth but the condition is increasingly being recognized at prenatal anomaly screening by US [38, 39]. There are several classifi- cation systems but none is linked to management protocols [40]. Treatment has been little changed for some time. Manipulation, splinting and often surgical soft tissue release are employed. For late problems, osteotomy and fusion are occasionally required [41]. 1.4.2 Role of Imaging Imaging is now often used to make an intrauterine diagnosis. For postnatal assessment some use MRI to assess the bony anatomy but the structures are very small and infants often will require anaesthesia for effective examination. Ossification of the hind foot bones is minimal in the infant where surgery is first considered. For this reason CT has little to offer, but plain films will help to clarify the overall alignment of the major bones. Plain radiographs are taken with an assistant holding a wooden block against the foot to achieve a “standing” position of the foot. The align- ment of the hind foot is most important; MR studies have shown abnormal rotation and equinus of the calcaneus [42, 43]. The axis of the talus should align with the first metatarsal and the axis of the calca- neus should align with the fourth or fifth metatar- sal (Fig. 1.9a). On a “standing” lateral view the talus should align with the first metatarsal whilst the cal- caneus should make an angle of 10–30° with the talus and align with the first metatarsal (Fig. 1.9b). The observer should first judge the alignment of the hind foot as varus, valgus or normal. Then the relative position of the forefoot on the frontal (a.p.) view may be assessed. Hind foot valgus usually leads to a com- pensatory forefoot varus. The talus may be normally aligned or in a vertical position. The most common malalignment of the calcaneus is into “equinus” posi- tion. Named after the position of the horse’s calca- neus, this implies an abnormal vertical alignment of the calcaneus with an excessively high arch to the mid-foot. US has the advantage of being dynamic and will assess the soft tissues in all but the most agitated of children [44]. It can also demonstrate the position of unossified bones [45]. As the aim of imaging is to define the abnormalities to allow planning of surgery, there must be close collaboration and understanding between the ultrasonographer and the surgeon. For this reason MR is probably the most useful technique [46]. MR imaging may be technically demanding as the deformity makes standard planes difficult to identify and reproduce. It is often easiest to strap the foot to a plastic or wooden splint to achieve as close to normal alignment as possible, this being equiva- lent to the walking position. The three conventional planes (coronal, sagittal and axial) are then used with sequences designed to contrast cartilage, muscle and tendon. T2-weighted fast spin echo is the most useful. Attention should be paid the number and alignment of the hind foot bones. Tibialis posterior tendon ten- sion is often implicated and it is helpful to identify this tendon. Despite the potential for demonstrating the static anatomy, many surgeons will rely on clinical examination and the response to manipulation under anaesthesia for their diagnosis, classification and assessment. Postoperative imaging is probably best achieved with MRI [47, 48] when the position of bone, unossified cartilage and tendons may be studied. There are links between lower limb deformity and spinal lesions so that careful clinical review of the spine with consideration of specific imaging is important in all children with foot deformities [49]. . Background 1 1 .2. 2 Role of Imaging in Detection 2 1 .2. 2.1 US Methods 3 1 .2. 3 Role of Imaging in Treatment 4 1 .2. 4 Potential Developments 5 1.3 Focal Defects 7 1.3.1 Clinical Background 7 1.3 .2 Role of. con- sensus that this is reasonable at present [7, 23 28 ]. Not least is the doubt that splint treatment is neces- sary in all abnormal cases [29 ]. National policies on screening will in part. used in high-risk screening protocols) First degree relative with hip dysplasia Premature birth Breech presentation Other congenital limb defects Spinal defects 1 .2. 2.1 US Methods 1 .2. 2.1.1 Morphology The